Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
Influence of Asymmetric Contact Form on Contact Resistance and Schottky Barrier, and Corresponding Applications of Diode Yudan Zhao, Xiaoyang Xiao, Yujia Huo, Yingcheng Wang, Tianfu Zhang, Kaili Jiang, Jiaping Wang, Shoushan Fan, and Qunqing Li ACS Appl. Mater. Interfaces, Just Accepted Manuscript • Publication Date (Web): 15 May 2017 Downloaded from http://pubs.acs.org on May 21, 2017
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 32
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
Influence of Asymmetric Contact Form on Contact Resistance and Schottky Barrier, and Corresponding Applications of Diode Yudan Zhao1,2, Xiaoyang Xiao1,2, Yujia Huo1,2, Yingcheng Wang1,2, Tianfu Zhang1,2, Kaili Jiang1,2, Jiaping Wang1,2, Shoushan Fan1,2, Qunqing Li1,2* 1
State Key Laboratory of Low-Dimensional Quantum Physics, Department of Physics and
Tsinghua-Foxconn Nanotechnology Research Center, Tsinghua University, Beijing 100084, China 2
Collaborative Innovation Center of Quantum Matter, Beijing 100084, China
*Corresponding author. E-mail address:
[email protected] ABSTRACT
We have fabricated carbon nanotube and MoS2 field effect transistors with asymmetric contact forms of source-drain electrodes, from which we found the current directionality of the devices and different contact resistances under the two current directions. By designing various structures, we can conclude that the asymmetric electrical performance was caused by the difference in the effective Schottky barrier height (ΦSB) caused by the different contact forms. A
ACS Paragon Plus Environment
1
ACS Applied Materials & Interfaces
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 32
detailed temperature-dependent study was used to extract and compare the ΦSB for both contact forms of CNT and MoS2 devices; we found that the ΦSB for the metal-on-semiconductor form was much lower than that of the semiconductor-on-metal form and is suitable for all p-type, ntype, or ambipolar semiconductors. This conclusion is meaningful with respect to the design and application of nanomaterial electronic devices. Additionally, using the difference in barrier height caused by the contact forms, we have also proposed and fabricated Schottky barrier diodes with a current ratio up to 104; rectifying circuits consisting of these diodes were able to work in a wide frequency range. This design avoided the use of complex chemical doping or heterojunction methods to achieve fundamental diodes that are relatively simple and use only a single material; these may be suitable for future application in nanoelectronic radio frequency or integrated circuits.
KEYWORDS carbon nanotube, MoS2, contact form, Schottky barrier, diode
INTRODUCTION Because of their unique electrical, optical, mechanical, and chemical properties, lowdimensional semiconducting materials have attracted tremendous interest and have been studied widely all over the world in the past few years. Among them, the one-dimensional semiconducting carbon nanotube (CNT) and two-dimensional few-layer molybdenum sulfide (MoS2) are two of the most attractive semiconducting materials and have been considered as candidates for the conductive channel of future field-effect transistors (FETs) (or thin film transistors (TFTs)) in nanoelectronic applications1-6 because of their great current capacity,7 high mobility,8-9 large on/off current ratio,10-11 and even outstanding mechanical performance.12-13
ACS Paragon Plus Environment
2
Page 3 of 32
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
However, there are still some obstacles that need to be solved before their real use and manufacture, one of which is the performance limitation of metal-semiconductor contacts,9, 14-15 including contact materials16-18 and contact forms.19-21 To now, the focus has concentrated on evaluating and studying metal-semiconductor contacts by comparing various electrode materials for both CNT and MoS2 devices, but not much attention has been paid to comparing the metalsemiconducting contact forms carefully, including semiconductor-on-metal (S-on-M) and metalon-semiconductor (M-on-S) or embedded forms, an area which is also decisive and worthy of discussion.22-23 Moreover, the relevant contrasting experiments with respect to contact forms are limited and the point of view is not uniform;14, 16 thus, a thorough experimental study on the contact forms is necessary. Different from conventional bulk silicon technology, it is difficult to realize controllable and stable doping in nanomaterials. Thus, the diode, an important basic electronic component in the modern semiconducting industry, cannot be achieved by controlled doping of a p-n junction in nanodevices, which is currently the most widely used technology for diodes. Recently, there have been some nanomaterial diodes fabricated using a heterojunction method,24-27 asymmetric electrode materials method,
28-34
and chemical doping method,
35-37
but a simpler fabrication
technology and even the use of a single material remain barriers to achieving diodes for future large-scale nanoelectronic manufacture. Therefore, to further study these two aspects, we fabricated one kind of asymmetric transistor based on either CNT and MoS2, in which the source and drain (S-D) electrode material is the same but their contact forms are different, e.g., the source is S-on-M and drain is M-on-S, in the same transistor. We found that the asymmetric contact forms could lead to an asymmetric electrical performance if we exchanged the position of S and D (the bias voltage position), which
ACS Paragon Plus Environment
3
ACS Applied Materials & Interfaces
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 32
means the device has directionality. Thus, to determine the influence of the contact forms on the electronic performance, the transfer length method (TLM) 14, 16, 38 and Y-function method (YFM) 39-41
were used to compare and calculate the contact resistance (Rc) of the same device under
different current directions; also, a detailed temperature-dependent study that considered both thermionic emission over the Schottky barrier as well as thermally assisted tunneling was used to extract the effective Schottky barrier height (ΦSB) 17, 42-44 of the two different contact forms using the same device. As a result, we found that the contact resistance showed a difference of approximately four times (5 kΩ and 20 kΩ) when the current flowed from the M-on-S to S-on-M electrode or vice versa, which denotes a different current direction for the asymmetric CNT-TFT; the effective Schottky barrier heights of CNT-on-metal and metal-on-CNT, as well as MoS2-onmetal and metal-on-MoS2, were 114 and 86 meV, and 95.4 and 32.4 meV, respectively. These results indicate that the contact form can influence the electrical performance to a large extent, and the Schottky barrier was concluded to be the determining mechanism, caused by different contact forms, by comparison of several device structures, no matter how long the contact length is or whether it is embedded or not. The conclusion that the ΦSB of the metal-on-semiconductor form is much lower than that of the semiconductor-on-metal form means that it is widely suitable for all p-type, n-type, or ambipolar semiconductors, which is helpful with respect to device design and fabrication. Meanwhile, taking notice of the different Schottky barrier heights under different contact forms, we also proposed and fabricated a Schottky barrier diode using the asymmetric contact form and applied it in half-wave and full-wave rectifying circuits. The current ratio of the diodes was up to approximately 104 under VDS ranges from -3 V to 3 V, and rectifying circuits could work normally under a wide input frequency range. This easy processing method can solve diode
ACS Paragon Plus Environment
4
Page 5 of 32
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
application problems in future nanoelectronic radio frequency or integrated circuits, avoiding the use of complex chemical doping or heterojunction methods.
RESULTS AND DISCUSSION We used a deposited semiconducting CNT network45 and few-layer MoS2 as conductive channel materials to study the influence of contact forms. Because the experimental phenomena associated with MoS2 fabricated by both chemical vapor deposition (CVD) growth and mechanical exfoliation are the same, here we only demonstrate the results of mechanically exfoliated MoS2. We first found the asymmetric phenomena of CNT-TFTs through the double-layer source and drain electrode structure with global back gate as shown in Figure 1, the thickness of thermal SiO2 dielectric layer is 300nm. Although the embedded contact form is almost the same for both S and D, which occupies most of the contact region, there is a large current difference in the same TFT under different current directions, as the red and blue transfer curves in Figure 1c show. We suggest that the unavoidable deviation of the two layers of electrodes caused by the error in standard photolithography, which leads to opposing contact forms at the edge of the S and D electrodes, causes the asymmetric electrical properties, as the schematic and scanning electron microscope image in Figure 1a and 1b show.
ACS Paragon Plus Environment
5
ACS Applied Materials & Interfaces
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 6 of 32
Figure 1. (a) Schematic and (b) scanning electron microscope images of the embedded doublelayer electrode CNT-TFT. An obvious deviation in the two layers can be seen. (c) Transfer curves under VDS= -1V for the same TFT under different current directions show a large difference. To verify the influence of the different contact form on the electrical performance of TFTs, we fabricated simpler and more particularly asymmetric TFTs with both CNT networks and MoS2, as shown in Figure 2 and Figure S1. Figure 2a and S1a show images of the asymmetric devices using a CNT network and MoS2 as channels, respectively, in which the bottom contact is S-on-M and the top is M-on-S. Normally, for a conventional symmetric contact form, as Figure 2d and S1c show, the transfer characteristics should be absolutely the same, no matter which side is selected as S (or D). However, for the asymmetric transistors shown in Figure 2c and S1b, the transfer curves show obvious differences when the source position is exchanged from one side to the other side (which means the current direction is exchanged). Schematics of CNT-TFTs and corresponding current direction in Figure 2b describe this phenomenon clearly. In this paper, we used VDS = -1 V for p-type CNT devices and VDS = 1 V for n-type MoS2 devices. To distinguish
ACS Paragon Plus Environment
6
Page 7 of 32
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
the two different current directions, we have labelled the larger current (I1) of both CNT and MoS2 devices with red curves and arrows and have used blue curves and arrows to indicate the smaller one (I2); similar notation is used throughout. While the corresponding calculation under the I1 direction are all red curves and that of I2 direction are all blue curves, for both CNT and MoS2 throughout. Comparing the CNT and MoS2 FETs, it is apparent that the larger current direction for p-type CNT FETs flows from the electrode of M-on-S to S-on-M, while that of the MoS2 devices is the opposite; this will be explained by the difference between electrons and holes later.
Figure 2. (a) SEM image, (b) schematics and (c) transfer and output characteristics of a typical local back-gate CNT-TFT with asymmetric contact forms. The drain current differs because of the direction. (d) Transfer and output characteristics of a typical local back-gate CNT-TFT with symmetric contact forms. No obvious directionality of the current is observed. Dashed lines here
ACS Paragon Plus Environment
7
ACS Applied Materials & Interfaces
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 8 of 32
represent retest curves after exchanging the position of two cables and probes. All VDS= -1 V here. Before analyzing the mechanism of the asymmetric electrical performance, the probable influence of probes and cables needs to be prevented. We therefore exchanged the two probe positions and tested the device again, achieving the same results, as the dashed lines in Figure 2 show (similar notation is also used throughout), which indicates that the asymmetric electrical performance is primarily caused by the structure, not instrument error. With the CNT-TFT as an example, we extracted the contact resistance of the same device under different directions using the Y-function method. Each direction of the transistors was also tested twice before and after exchanging the position of probes and cables, as the filled dotted lines and hollow dotted lines in Figure 3 show. From the experimental data IDS-VGS curve, we can achieve the gm-VGS curve where gm=dIDS/dVGS. Then we define the y-function which can also be achieved through the tested IDS-VGS curve as: 39 Y=
= ( − )
(1)
where Gm= (µFE CG)/Lch2 is defined as the transconductance parameter, which can be calculated by the slope k of the linear region of the y-function, as Gm=k2/VDS, and threshold voltage VT can be extracted from the VGS-axis intercept of liner region of y-function. Finally, using the formula: 39
2, = − =
! #$ % (% &' )
−"
=
( ( % &' )
−
(2)
we can calculate Rc using the y-function, which can be achieved only using the transfer characteristic curves, and avoid measuring specific parameters of the transistors. From the yfunction in Figure 3a, we find that the slopes of the linear region of the y-functions are different,
ACS Paragon Plus Environment
8
Page 9 of 32
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
leading to a difference in Rc of approximately four times for these two directions. Similarly, we also define the contact resistance of the I1 direction as Rc1 and that of I2 as Rc2. Figure 3b shows that Rc1 is approximately 5 kΩ, while Rc2 is approximately 20 kΩ under the on state; and Rc1 possesses less than 80%, while Rc2 possesses greater than 95%, of the total device resistance (Rtot) under the on state. On the other hand, no distinction can be found in symmetric devices, as the yfunction and Rc in Figure 3c and 3d show.
Figure 3. Contact resistance calculated by the Y-function method. (a) Transfer characteristics and corresponding y-function of a typical CNT-TFT with asymmetric contact. (b) Contact resistance and the percentage of 2Rc/Rtot of the asymmetric device under two different directions. A difference in Rc of approximately four times and a 20 percent difference in 2Rc/Rtot reveal the tremendous influence of the contact forms. (c) Transfer characteristics and y-function and (d) Rc
ACS Paragon Plus Environment
9
ACS Applied Materials & Interfaces
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 32
and 2Rc/Rtot of a typical symmetric contact CNT-TFT. There is absolutely no difference in electrical performance between the two directions. Dashed lines here also represent retest curves after exchanging the cables and probes. All VDS= -1 V for transfer curves here. The results indicate an enormous influence on Rc by the contact forms. Although the current flows through both source and drain electrodes no matter the direction, the Rc we calculated still confirms that the current direction is decisive, so the asymmetric contact form affects the electrical performance greatly. However, the Y-function method can only determine the total contact resistance; it cannot distinguish which side plays a leading role. Additionally, it is doubt that if the current difference is judged by the contact resistance, the current would be different under every gate voltage. However, the phenomenon mainly reveal near the on-state, not in the off-state. Moreover, it is also important to understand the mechanism by which the contact form can influence the electrical performance of FETs. Thus, according to the basic experimental results above, a unilateral suspended semiconducting channel that cannot be modulated;46-47 a main-area contact difference caused by S-on-M or M-on-S in the metal and semiconductor contact region;16, 38, 48 and a Schottky barrier difference near the metal-semiconducting contact edge9, 49-51 are the three possible mechanisms that may cause the directionality of devices. Therefore, we designed CNT-TFTs with six structures to determine the mechanism behind the asymmetric electrical performance, as Figure 4 and Figure S5 in Supporting Information shows. In each structure, we demonstrate a schematic of the structure, transfer curves of both current directions, and the change in resistance/current with various channel lengths under the on state (the TLM in Figure S5). Because the contact resistance of CNT network TFTs cannot be extracted by TLM accurately,16 we only use it to compare the Rc qualitatively.
ACS Paragon Plus Environment
10
Page 11 of 32
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
Figure 4. Comparison of transfer characteristics for six different structures (VDS= -1 V here). Symmetric contact forms with (a) both M-on-S and (b) both S-on-M do not show current
ACS Paragon Plus Environment
11
ACS Applied Materials & Interfaces
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 12 of 32
directionality. (c) When both contact forms are S-on-M but one side of the channel is suspended, there is no electrical difference. (d) When the contact forms are different but have no unilateral suspended channel, asymmetric electrical performance is exhibited, which indicates that contact form is the dominant cause, not the unilateral suspended channel. (e) Typical asymmetric contact forms show distinctions in the on current. (f) A double-layer S-D electrode with a tiny deviation shows electrical asymmetry, proving the mechanism caused by contact forms is due to the Schottky barrier of the semiconductor-metal contact edge, not the difference in the overlapping main contacting area. The SiO2 dielectric in c, d and f was etched by CF4 to bury the electrodes, and the depth of which is 50 nm. From Figure 4a and 4b, it is clear that the transfer curves of the two directions coincide perfectly under symmetric structures; there is also no distinction between the intercepts of the resistance-channel length lines, which suggests that there is no difference in 2Rc, as Figure S5a and S5b shows. The structure in Figure 4c has only a unilateral suspended CNT channel with both S-on-M contact forms (the other one is a buried electrode), but it does not show asymmetric electrical performance, as do structure 1 and 2. In contrast, structure 4 in Figure 4d has no suspended channel but does have an S-on-M and an M-on-S contact, showing asymmetric transfer curves and Rc as we demonstrate above, as does the original structure 5 in Figure 4e. Therefore, we can conclude that the main mechanism that influences the electrical performance and leads to asymmetric FETs is not the suspension of the channels. Some studies have shown that the overlapping conducting region of metals and semiconductors could influence the performance of TFTs to a large extent because of the conductive length
15, 38
or van der Waals
gap.22-23 Thus, we analyzed the CNT-TFT of structure 6 with a double-layer S-D as an embedded contact form, as Figure 4f shows, which is the same as that in Figure 1. In this structure, the main
ACS Paragon Plus Environment
12
Page 13 of 32
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
contact form of both S and D is almost the same and symmetric (M-on-S-on-M), as well as the contact resistance, contact length, or gap in the main contact region. However, because of unavoidable error in the photolithography process, the two layer electrodes may have a tiny deviation, of approximately dozens or hundreds of nanometers, leading to a totally opposite contact form at the metal-semiconductor contact edge. Thus, the existence of asymmetric electrical performance in structure 6 reveals that the metal-semiconducting contact form near the asymmetric edge is the main mechanism, which means that the different contact Schottky barrier height here plays a leading role, not the main-area contact forms, even though their area is much far larger than that of the asymmetric edges. Detailed images and analysis can be found in section 1 of the Supporting Information. It is also need to say that the phenomenon of asymmetric current mainly reveal near the on-state, not in the off-state. Thus we believe the Schottky barrier is the real mechanism because the Schottky barrier becomes dominated only near on-state (when VGS over the flat band, tunneling current plays the leading role), while Schottky barrier is not dominated near the off-state (when VGS below the flat band, thermionic current plays the leading role). Next, since the Schottky barrier difference is suggested to be the main mechanism because of the contact forms, an accurate experiment based on the thermal emission and tunneling current theory dependence on temperature52 for Schottky barrier devices is carried out to extract the effective Schottky barrier height for different contact forms.
ACS Paragon Plus Environment
13
ACS Applied Materials & Interfaces
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 14 of 32
Figure 5. Transfer characteristics of a typical (a) CNT-TFT and (d) exfoliated MoS2 FET with different current directions (labeled by solid lines and dashed lines) under various temperatures from 100 K to 400 K. ln (Id/T3/2) versus 1000/T under different gate biases extracted from transfer curves of (b) CNT-TFT and (e) MoS2-FET in both directions. The slope of dots with the same gate bias (in the same color) in the high-temperature area can be used to calculate the barrier height (ΦB) under specific gate voltages. ΦB for different gate biases of (c) CNT-TFT and (f) MoS2-FET in the two current directions, showing different effective Schottky barrier heights and corresponding flat-band voltages. VDS= -1 V for CNT-TFTs and VDS= 1 V for MoS2 FETs here. Figure 5a and 5d show the transfer characteristics of a typical CNT-TFT and exfoliated MoS2 FET with different current directions under temperatures ranging from 100 K to 400 K. According to the thermionic theory, when the gate voltage (VGS) is below the flat-band voltage (VFB), the drain current can be written as:
ACS Paragon Plus Environment
14
Page 15 of 32
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
)* = +∗,* . //, exp 4
567 87
9 [1 − exp (−
5=> 87
)]
(3)
where +∗,* is the 2D equivalent Richardson constant (we consider the CNT network to be twodimensional as well), T is the absolute temperature, kB is the Boltzmann constant, q is the electron charge, and Vds is the drain to source bias, i.e., 1 V for n-type MoS2 and -1 V for p-type CNT transistors.17, 43, 53 Therefore, the value of ln (Id/T3/2) under different temperatures can be drawn from the transfer curves under each gate bias voltage. Figure 5b and 5e show the Arrhenius plot of ln (Id/T3/2) versus 1000/T under different gate biases for CNT and MoS2 transistors, respectively, and the barrier height under specific gate biases can be calculated by the slope of each line (dots in the same color) in the high-temperature region, in which the round dots represent ln (Id/T3/2) with the current direction I1, and stars represent that of I2. Using the slopes of the high-temperature region we extracted, graphs of ΦB versus gate bias voltage are drawn in Figure 5c and 5f for both CNT and MoS2 transistors in both directions. For typical Schottky barrier devices, when VGS is below VFB, thermionic emission current plays the dominant role, while when VGS becomes larger than VFB, thermally assisted tunneling current becomes increasingly significant, resulting in nonlinear behavior. Therefore, ΦSB can be most accurately extracted from the end of the linear region, where VGS = VFB.17, 43 After that, the on current is mainly decided by the tunneling current which is determined by the ΦSB because of the pining of Schottky barrier, and the ΦB of thermionic current should not be taken into consideration. In Figure 5c, the red dotted lines are extracted from the transfer curve for the I1 direction, which means that drain bias voltage (VD = -1 V) is added to the CNT-on-metal electrode, while the blue dotted lines are extracted from that for the I2 direction, where bias voltage is added to the metal-on-CNT electrode. Because the Schottky barrier primarily acts on the other source
ACS Paragon Plus Environment
15
ACS Applied Materials & Interfaces
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 16 of 32
electrode side (VS = 0 V), as the inner schematic of Figure 5c shows, we can obtain ΦSB = 86 meV for the metal-on-CNT contact and ΦSB = 114 meV for the CNT-on-metal contact. The height of ΦSB exactly corresponds to the direction of the on current and Rc. Because the sweep direction of the local back-gated VGS for the CNT-TFT here was from -3 V to 3 V, the tunneling current dominates at first for both types, located in the white area of Figure 5c. As the gate bias moves in the positive direction, the energy band bends downwards to the same extent for both directions,54 but because of the lower ΦSB for the metal-on-CNT side, VGS could arrive at VFB first, obstructing the tunneling current in the light-gray area in Figure 5c. For the blue type, because of the higher ΦSB for the CNT-on-metal side, the VGS here cannot bend the energy band to the flat band, thus tunneling current is still allowed in this area. As the VGS moves in the positive direction enough for the VFB of the CNT-on-metal side, thermal emission begins to play the principal role for both of them, as shown in the dark-gray area in Figure 5c. In summary, gate voltage here plays an important role of modulating the energy band as well as effective Schottky barrier to decide which kind of current takes the lead, and thus we can find the difference of drain current under the two directions is relevant to the gate voltage closely as a result. Similarly, Figure 5f shows ΦSB = 32.4 meV for the metal-on-MoS2 contact and ΦSB = 95.4 meV for the MoS2-on-metal contact. The height of ΦSB here also exactly corresponds to the direction of the on current and Rc. The global back-gate sweep direction of VGS is from -50 V to 50 V, which means the energy band bends upwards at first and only thermal emission current can flow through, as seen in the dark-gray area in Figure 5f. As VGS moves in the positive direction, the energy band bends downwards and first meets the higher flat-band of the MoS2-on-metal contact (blue type). Since then tunneling current begins to flow in the blue type while it is still blocked in the red type, as shown in the light-gray area of Figure 5f. Finally, after VGS moves to
ACS Paragon Plus Environment
16
Page 17 of 32
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
the VFB of the red type, both of the two types are primarily dominated by the tunneling current, arriving in the white area in Figure 5f. The mechanism that the metal-semiconductor junction configuration influence the effective SBH has been studied previously. In some recent reports, calculations of the interface interaction between the metal electrode and semiconductor have proved that junction geometry55,
56
can
influence the SBH to a certain extent; and the substrate inhomogeneity which semiconductor located on (SiO2 for M-on-S side and metal for S-on-M side here) can also affect the SBH. 57-59 Therefore, it is likely that the different metal-semiconductor junction configuration of the asymmetric contact transistor here lead to the different SBH. Further on, we believe the difference of van der Waals gap22 between semiconductor and metal caused by their relative location could lead to different effective Schottky barrier heights, which is caused by the interaction between metal-semiconductor and substrate inhomogeneity mentioned above. Normally, the Schottky barrier height is only judged by the work functions of metal and semiconductor, but here the different interaction between them caused by different junction geometry can lead to different effective SBH. The experiments and analysis illustrated above confirm that the effective Schottky barrier height of the S-on-M contact form is higher than that of the M-on-S for both n-type and p-type devices. Moreover, as we proposed above in Figure 2 and Figure S1, the phenomenon in which I1 and I2 of the CNT-TFT and MoS2-FET show opposing properties can be explained now because electrons in n-type devices and holes in p-type devices have similar transmission properties, while the same direction of electron transport and hole transport could lead to a converse current direction. This is why the larger current of CNT-TFT flows from M-on-S to Son-M, while that of MoS2-FET flows in the opposite direction.
ACS Paragon Plus Environment
17
ACS Applied Materials & Interfaces
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 18 of 32
Furthermore, asymmetric contact form FETs with CNT aligned arrays and single CNTs as conductive channels with global back-gates were also fabricated to verify the experiments and theory we proposed above, as Figure S4 in section 3 of the Supporting Information demonstrates. Because the FETs show ambipolar properties, both electrons and holes could be transported in channels under the two branches. The result is that both of the branches confirm the current phenomena of p-type and n-type devices, which indicates that the theory of the Schottky barrier difference caused by the contact forms is correct and widely suitable, even for ambipolar devices. On the other hand, taking advantage of the difference in Schottky barrier because of the asymmetric contact transistors, we can fabricate a Schottky barrier diode using this property, as the schematic in Figure 6b shows; and the global back-gate bias can also be added or not. To amplify the effect of the barriers, we minimize the channel length down to 1~2 µm and add a global back-gate voltage to turn all CNTs and MoS2 to the on state, achieving a highperformance Schottky diode, as shown in Figure 6a. Diodes without back-gate bias can also work well, which is shown in section 5 of the Supporting Information. Key electrical performance and fabrication processes of several recently reported low-dimensional semiconductor diodes are compared in Table 1, from which we can see that this method for highperformance diode fabrication is simpler and only involves a single conductive material, avoiding complex chemical doping or heterojunction methods.
ACS Paragon Plus Environment
18
Page 19 of 32
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
Figure 6. Schottky barrier diode achieved by asymmetric contact forms: (a) Transfer curves under two directions (VDS= -1 V) and corresponding I-V curve of MoS2 diode; the current ratio could be up to 104 when sweep range is from -3 V to 3 V under VGS=50 V. (b) MoS2 diode can work normally when the bias voltage range decreases to -0.5 V to 0.5 V under VGS=50 V and can keep a current ratio near 103. Inner schematic: Structure and direction of the MoS2 diode, the global back-gate bias can both be added or not. (c) CNT network diode with opposite bias voltage, showing inverse I-V curves, which means the diode is very reliable. The schematic and input-output waveforms of (d)-(e) half-wave rectifying circuits and (f)-(g) full-wave rectifying circuits consist of asymmetric contact diodes. Output voltages are both attenuated to some extent but not distorted. The small deviation between the two periods in full-wave rectifying circuits may be due to the slight inconformity of the diodes. Load resistance is fabricated and adjusted by cutting CNT thin films into strips.
ACS Paragon Plus Environment
19
ACS Applied Materials & Interfaces
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 20 of 32
Figure 6a and 6c show the Schottky diode achieved by asymmetric contact forms with backgate bias, using MoS2 and CNT as conductive channels, respectively. The on current difference in the two directions could be up to approximately 104, so the current ratio is 104 under VDS from -3 V to 3 V, as Figure 6a shows. In Table 1, we can also find that the maximum current ratio and ON current of our work are located in the top level among the recently reported diodes based on low-dimensional semiconductors, while the fabrication processes are able to be large-scale and no need of chemical doping, electron-beam lithography or site-directed transfer, only with one kind of semiconductor and metal. The current ratio would decrease if the range of VDS was reduced to 0.5 V, but it can still remain over 103, as Figure 6b shows. The asymmetric diodes can still work with current ratio between 102~103 without back-gate voltage as Figure S6 shows, which gives designers various choices to meet the specific needs. From Figure 6c, inverse I-V curves of the CNT-based diode could be measured if we change the current direction, which means the direction of the diode is fixed. Additionally, we apply the diodes we designed into half-wave and full-wave rectifying circuits,60 as Figure 6d to 6g show. Load resistance here is fabricated and adjusted by cutting CNT thin films into strips to meet our requirements. As a result, although the output waveforms are not distorted, they still have some attenuations (2 V down to 1.65 V for half-wave and 2 V down to 1.35 V for full-wave); the full-wave output also has a small deviation between the two periods of 0 to π and π to 2π. This could be caused by the circuit design and slight inconformity of our diodes. However, the diodes can still work normally under a wide range of frequency, from 0.1 kHz to 50 kHz, which is an important foundation for primary element applications in future Carbon nanotube (or 2D-material) based nanoelectronic6163, 66
integrated or radio frequency circuits.
ACS Paragon Plus Environment
20
Page 21 of 32
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
Table 1. Performance and fabrication processes comparison of diodes based on lowdimensional semiconductors.
Type of diodes
Maximum Current ratio
Ion(A)
Asymmetric contact forms (CNT and MoS2)
>104
~2x10-6
✔
MoS2 heterojunction diodes 24, 25, 65
10~103
~10-6
Asymmetric S/D metals (CNT)28-31, 34
102~104
Split-gate (CNT)32
No Largechemical scale doping
No electronbeam lithography
No Sitedirected transfer
✔
✔
✔(CNT)
✘
✔
✘
✘
~1x10-6
✘
✔
✘
✔
~107
2x10-8
✘
✔
✘
✔
Pt-Ti asymmetric S/D (MoS2)33
102
2x10-6
✘
✔
✘
✔
CONCLUSIONS In summary, we have systematically studied the influence of contact forms on electrical performance through asymmetric field effect transistors based on CNTs and MoS2, including transfer characteristics, the on current, contact resistance, and the Schottky barrier. By comparing various structures, we found the real mechanism of the contact forms was the difference in effective Schottky barrier height caused by the metal and semiconductor contact edge. Arrhenius plots were drawn through temperature-dependent studies to extract the ΦSB, and it was found that for all p-type, n-type, or ambipolar semiconductors, the ΦSB of the metal-on-semiconductor form
ACS Paragon Plus Environment
21
ACS Applied Materials & Interfaces
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 22 of 32
was much lower than that of the semiconductor-on-metal form, leading to the directionality of asymmetric devices and opposing properties between transport phenomena of electrons and holes. Finally, taking advantage of the different barrier heights, we proposed and fabricated a Schottky barrier diode using the asymmetric structure, the current ratio of which could be as high as 104. Rectifying circuits consisting of the diodes could work normally under a wide frequency range, indicating the diodes are reliable for future applications of nanoelectronic integrated circuits. This method is simple and only requires a single material, avoiding the complex chemical doping and heterojunction methods, to achieve high-performance diodes.
METHODS Deposition of CNT thin films. Semiconducting SWCNT suspensions in methylbenzene (more than 99.9% pure) were prepared by pre-separation of arc-discharged SWCNTs. Before deposition, the substrates were treated via O2 plasma etching to functionalize the oxide surfaces and make them hydrophilic. Then, the samples were immersed in the semiconducting nanotube suspension for several hours to deposit the SWCNT random network thin films on the substrate. Finally, the substrate was refloated and dried using N2. CNT aligned arrays and single-CNT channels were grown by CVD and then transferred onto target electrodes. Preparation of MoS2. The MoS2 channels that we used were prepared by CVD growth and mechanical exfoliation. For the CVD method, we deposited MoS2 on SiO2/Si substrates using sulfur and MoO3 powder as the precursors. The typical growth temperature was 850 °C, and argon was used as the carrier gas. For mechanical exfoliation, we exfoliated a MoS2 sample with several layers from a MoS2 mineral.
ACS Paragon Plus Environment
22
Page 23 of 32
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
Fabrication of devices. Electrodes, test holes, and conductive channels were all defined using standard photolithography. Electrodes for the CNT-TFTs were 50-nm-thick Pd layers, and the conductive channels were patterned using O2 plasma etching. Electrodes for the MoS2 FETs were 50-nm-thick Au layers. It is also need to say that the same results achieved when using Pd electrodes for MoS2 FETs while using Au electrodes for CNT TFTs. The asymmetric channels of MoS2 were fabricated using a site-specific method to transfer few-layer MoS2. All buried electrodes here was fabricated by filling the SiO2 holes which is etched by CF4 reacting ion etching under the condition of 40 sccm, 2 Pa, 50 W, and the speed is about 1 nm/ 1.6 s. Characterization of electrical properties. All electrical measurements were performed using a semiconductor analyzer (4156C, Agilent, USA) with a probe station (TTP, Lakeshore, USA) under specific conditions. The output waves of the rectifying circuits were captured using a digital oscilloscope (TDS2012C, Tektronix, USA).
ASSOCIATED CONTENT Supporting Information. The Supporting Information is available online. Supporting figures and discussions contain (1) MoS2 field-effect transistor (FET) with asymmetric contact forms, (2) carbon nanotube (CNT) thin film transistor (TFT) with double-layer source and drain electrode (embedded contact), (3) ambipolar aligned-array CNT-TFT and single-CNT transistor with asymmetric S-D contact forms, (4) comparison of Ion and resistance with channel length (TLM) for the six structures, (5) electrical performance of MoS2 Schottky barrier diode without back-gate voltage, (6) Characterization of exfoliated and CVD growth MoS2, and (7) comparison of transfer characteristics of MoS2 FETs with different structures.
ACS Paragon Plus Environment
23
ACS Applied Materials & Interfaces
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 24 of 32
ACKNOWLEDGMENTS This work was financially supported by the National Basic Research Program of China (2012CB932301) and the National Natural Science Foundation of China (51532008, 11574171).
REFERENCES (1) Hu, L.; Hecht, D. S.; Gruner, G. Carbon Nanotube Thin Films: Fabrication, Properties, and Applications. Chem. Rev. 2010, 110, 5790–5844. (2) Zhao, Y.; Li, Q.; Xiao, X.; Li, G.; Jin, Y.; Jiang, K.; Wang, J.; Fan, S. Three-Dimensional Flexible Complementary Metal-Oxide-Semiconductor Logic Circuits Based on Two–Layer Stacks of Single-Walled Carbon Nanotube Networks. ACS Nano 2016, 10, 2193−2202. (3) Zou, Y.; Li, Q.; Liu, J.; Jin, Y.; Qian, Q.; Jiang, K.; Fan, S. Fabrication of All-Carbon Nanotube Electronic Devices on Flexible Substrates through CVD and Transfer Methods. Adv. Mater. 2013, 25, 6050–6056. (4) Zhang, J.; Wang, C.; Zhou, C. Rigid/Flexible Transparent Electronics Based on Separated Carbon Nanotube Thin-Film Transistors and Their Application in Display Electronics. ACS Nano 2012, 6, 7412–7419. (5) Sun, D. M.; Timmermans, M. Y.; Tian, Y.; Nasibulin, A. G.; Kauppinen, E. I.; Kishimoto, S.; Mizutani, T.; Ohno, Y. Flexible High-Performance Carbon Nanotube Integrated Circuits. Nat. Nanotechnol. 2011, 6, 156–161. (6) Zhang, E.; Wang, W.; Zhang, C.; Jin, Y.; Zhu, G.; Sun, Q.; Zhang, D. W.; Zhou, P.; Xiu, F. Tunable Charge-Trap Memory Based on Few-Layer MoS2. ACS Nano 2015, 9, 612–619.
ACS Paragon Plus Environment
24
Page 25 of 32
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
(7) Yao, Z.; Kane, C. L.; Dekker, C. High-Field Electrical Transport in Single-Wall Carbon Nanotubes. Phys. Rev. Lett. 2000, 84, 2941–2944. (8) Fuhrer, M. S.; Kim, B. M.; Durkop, T.; Brintlinger, T. High-Mobility Nanotube Transistor Memory. Nano Lett. 2002, 2, 755–759. (9) Das, S.; Chen, H.; Penumatcha, A. V.; Appenzeller, J. High Performance Multilayer MoS2 Transistors with Scandium Contacts. Nano Lett. 2013, 13, 100–105. (10) Derenskyi, V.; Gomulya, W.; Rios, J. M. S.; Fritsch, M.; Froehlich, N.; Jung, S.; Allard, S.; Bisri, S. Z.; Gordiichuk, P.; Herrmann, A.; Scherf, U.; Loi, M. A. Carbon Nanotube Network Ambipolar Field-Effect Transistors with 10(8) On/Off Ratio. Adv. Mater. 2014, 26, 5969– 5975. (11) Radisavljevic, B.; Radenovic, A.; Brivio, J.; Giacometti, V.; Kis, A. Single-Layer MoS2 Transistors. Nat. Nanotechnol. 2011, 6, 147–150. (12) Bozovic, D.; Bockrath, M.; Hafner, J. H.; Lieber, C. M.; Park, H.; Tinkham, M. Plastic Deformations in Mechanically Strained Single-Walled Carbon Nanotubes. Phys. Rev. B 2003, 67, 033407. (13) Pu, J.; Yomogida, Y.; Liu, K. K.; Li, L. J.; Iwasa, Y.; Takenobu, T. Highly Flexible MoS2 Thin-Film Transistors with Ion Gel Dielectrics. Nano Lett. 2012, 12, 4013–4017. (14) Cao, Q.; Han, S.; Tersoff, J.; Franklin, A. D.; Zhu, Y.; Zhang, Z.; Tulevski, G. S.; Tang, J.; Haensch, W. End-Bonded Contacts for Carbon Nanotube Transistors with Low, SizeIndependent Resistance. Science 2015, 350, 68–72. (15) Franklin, A. D.; Luisier, M.; Han, S.; Tulevski, G.; Breslin, C. M.; Gignac, L.; Lundstrom, M. S.; Haensch, W. Sub-10 nm Carbon Nanotube Transistor. Nano Lett. 2012, 12, 758–762. (16) Xia, J.; Dong, G.; Tian, B.; Yan, Q.; Zhang, H.; Liang, X.; Peng, L. Metal Contact Effect on
ACS Paragon Plus Environment
25
ACS Applied Materials & Interfaces
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 26 of 32
the Performance and Scaling Behavior of Carbon Nanotube Thin Film Transistors. Nanoscale 2016, 8, 9988–9996. (17) Das, S.; Chen, H.; Penumatcha, A. V.; Appenzeller, J. High Performance Multilayer MoS2 Transistors with Scandium Contacts. Nano Lett. 2013, 13, 100–105. (18) Javey, A.; Guo, J.; Farmer, D. B.; Wang, Q.; Wang, D.; Gordon, R. G.; Lundstrom, M.; Dai, H. Carbon Nanotube Field-Effect Transistors with Integrated Ohmic Contacts and High-κ Gate Dielectrics. Nano Lett. 2004, 4, 447–450. (19) Bai, P.; Li, E.; Lam, K. T.; Kurniawan, O.; Koh, W. S. Carbon Nanotube Schottky Diode: An Atomic Perspective. Nanotechnology 2008, 19, 115203. (20) Léonard, F.; Tersoff, J. Role of Fermi-Level Pinning in Nanotube Schottky Diodes. Phys. Rev. Lett. 2000, 84, 4693. (21) Knoch, J.; Appenzeller, J. Tunneling Phenomena in Carbon Nanotube Field-Effect Transistors. Physica Status Solidi (a) 2008, 205, 679–694. (22) Allain, A.; Kang, J.; Banerjee, K.; Kis, A. Electrical Contacts to Two-Dimensional Semiconductors. Nat. Mater. 2015, 14, 1195–1205. (23) Xia, F.; Perebeinos, V.; Lin, Y.; Wu, Y.; Avouris, P. The Origins and Limits of Metal– Graphene Junction Resistance. Nat. Nanotechnol. 2011, 6, 179–184. (24) Yu, J. H.; Lee, H. R.; Hong, S. S.; Kong, D.; Lee, H. W.; Wang, H.; Xiong, F.; Wang, S.; Cui, Y. Vertical Heterostructure of Two-Dimensional MoS2 and WSe2 with Vertically Aligned Layers. Nano Lett. 2015, 15, 1031–1035. (25) Cheng, R.; Li, D.; Zhou, H.; Wang, C.; Yin, A.; Jiang, S.; Liu, Y.; Chen, Y.; Huang, Y.; Duan, X. Electroluminescence and Photocurrent Generation from Atomically Sharp WSe2/MoS2 Heterojunction P-N Diodes. Nano Lett. 2014, 14, 5590–5595.
ACS Paragon Plus Environment
26
Page 27 of 32
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
(26) Xu, Y.; Cheng, C.; Du, S.; Yang, J.; Yu, B.; Luo, J.; Yin, W.; Li, E.; Dong, S.; Ye, P.; Duan, X. Contacts between Two- and Three-Dimensional Materials: Ohmic, Schottky, and P-N Heterojunctions. ACS Nano 2016, 10, 4895–4919. (27) Jariwala, D.; Sangwan, V. K.; Seo, J. T.; Xu, W.; Smith, J.; Kim, C. H.; Lauhon, L. J.; Marks, T. J.; Hersam, M. C. Large-Area, Low-Voltage, Antiambipolar Heterojunctions from Solution-Processed Semiconductors. Nano Lett. 2015, 15, 416–421. (28) Manohara, H. M.; Wong, E. W.; Schlecht, E.; Hunt, B. D.; Siegel, P. H. Carbon Nanotube Schottky Diodes Using Ti−Schottky and Pt−Ohmic Contacts for High Frequency Applications. Nano Lett. 2005, 5, 1469–1474. (29) Lu, C.; An, L.; Fu, Q.; Liu, J.; Zhang, H.; Murduck, J. Schottky Diodes from Asymmetric Metal-Nanotube Contacts. Appl. Phys. Lett. 2006, 88, 133501. (30) Cobas, E.; Fuhrer, M. S. Microwave Rectification by a Carbon Nanotube Schottky Diode. Appl. Phys. Lett. 2008, 93, 43120. (31) Li, H.; Zhang, Q.; Marzari, N. Unique Carbon-Nanotube Field-Effect Transistors with Asymmetric Source and Drain Contacts. Nano Lett. 2008, 8, 64–68. (32) Hughes, M. A.; Homewood, K. P.; Curry, R. J.; Ohno, Y.; Mizutani, T. An Ultra-Low Leakage Current Single Carbon Nanotube Diode with Split-Gate and Asymmetric Contact Geometry. Appl. Phys. Lett. 2013, 103, 133508. (33) Yoon, H. S.; Joe, H.; Jun Kim, S.; Lee, H. S.; Im, S.; Min, B.; Jun, S. C. Layer Dependence and Gas Molecule Absorption Property in MoS2 Schottky Diode with Asymmetric Metal Contacts. Scientific Reports 2015, 5, 10440. (34) Wang, S.; Zhang, Z.; Ding, L.; Liang, X.; Shen, J.; Xu, H.; Chen, Q.; Cui, R.; Li, Y.; Peng, L. A Doping-Free Carbon Nanotube CMOS Inverter-Based Bipolar Diode and Ambipolar
ACS Paragon Plus Environment
27
ACS Applied Materials & Interfaces
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 28 of 32
Transistor. Adv. Mater. 2008, 20, 3258–3262. (35) Huang, L.; Chor, E. F.; Wu, Y.; Guo, Z. Fabrication of Single-Walled Carbon Nanotube Schottky Diode with Gold Contacts Modified by Asymmetric Thiolate Molecules. Carbon 2010, 48, 1298–1304. (36) Chen, C.; Liao, C.; Wei, L.; Zhong, H.; He, R.; Liu, Q.; Liu, X.; Lai, Y.; Song, C.; Jin, T.; Zhang, Y. Carbon Nanotube Intramolecular P-I-N Junction Diodes with Symmetric and Asymmetric Contacts. Scientific Reports 2016, 6, 22203. (37) Park, H.; Jung, W.; Kang, D.; Jeon, J.; Yoo, G.; Park, Y.; Lee, J.; Jang, Y. H.; Lee, J.; Park, S.; Yu, H.; Shin, B.; Lee, S.; Park, J. Extremely Low Contact Resistance on Graphene through N-Type Doping and Edge Contact Design. Adv. Mat. 2016, 28, 864–870. (38) Franklin, A. D.; Chen, Z. Length Scaling of Carbon Nanotube Transistors. Nat. Nanotechnol. 2010, 5, 858–862. (39) Cao, Q.; Han, S.; Tulevski, G. S.; Franklin, A. D.; Haensch, W. Evaluation of Field-Effect Mobility and Contact Resistance of Transistors That Use Solution-Processed Single-Walled Carbon Nanotubes. ACS Nano 2012, 6, 6471–6477. (40) Choi, S.; Bennett, P.; Takei, K.; Wang, C.; Lo, C. C.; Javey, A.; Bokor, J. Short-Channel Transistors Constructed with Solution-Processed Carbon Nanotubes. ACS Nano 2013, 7, 798–803. (41) Brady, G. J.; Joo, Y.; Wu, M.; Shea, M. J.; Gopalan, P.; Arnold, M. S. Polyfluorene-Sorted, Carbon Nanotube Array Field-Effect Transistors with Increased Current Density and High On/Off Ratio. ACS Nano 2014, 8, 11614–11621. (42) Chen, J.; Odenthal, P. M.; Swartz, A. G.; Floyd, G. C.; Wen, H.; Luo, K. Y.; Kawakami, R. K. Control of Schottky Barriers in Single Layer MoS2 Transistors with Ferromagnetic
ACS Paragon Plus Environment
28
Page 29 of 32
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
Contacts. Nano Lett. 2013, 13, 3106–3110. (43) Wang, J.; Yao, Q.; Huang, C.; Zou, X.; Liao, L.; Chen, S.; Fan, Z.; Zhang, K.; Wu, W.; Xiao, X.; Jiang, C.; Wu, W. High Mobility MoS2 Transistor with Low Schottky Barrier Contact by Using Atomic Thick h-BN as a Tunneling Layer. Adv. Mater. 2016, 28, 8302–8308. (44) Appenzeller, J.; Radosavljevic, M.; Knoch, J.; Avouris, P. Tunneling versus Thermionic Emission in One-Dimensional Semiconductors. Phys. Rev. Lett. 2004, 92, 048301. (45) Gu, J.; Han, J.; Liu, D.; Yu, X.; Kang, L.; Qiu, S.; Jin, H.; Li, H.; Li, Q.; Zhang, J. SolutionProcessable High-Purity Semiconducting SWCNTs for Large-Area Fabrication of HighPerformance Thin-Film Transistors. Small 2016, 12, 4993–4999. (46) Lin, Y.; Appenzeller, J.; Avouris, P. Ambipolar-to-Unipolar Conversion of Carbon Nanotube Transistors by Gate Structure Engineering. Nano Lett. 2004, 4, 947–950. (47) Li, J.; Zhang, Q. Simulation of Ambipolar-to-Unipolar Conversion of Carbon Nanotube Based Field Effect Transistors. Nanotechnology 2005, 16, 1415–1418. (48) Cummings, A. W.; Léonard, F. Enhanced Performance of Short-Channel Carbon Nanotube Field-Effect Transistors Due to Gate-Modulated Electrical Contacts. ACS Nano 2012, 6, 4494–4499. (49) Mann, D.; Javey, A.; Kong, J.; Wang, Q.; Dai, H. J. Ballistic Transport in Metallic Nanotubes with Reliable Pd Ohmic Contacts. Nano Lett. 2003, 3, 1541–1544. (50) Song, X.; Han, X.; Fu, Q.; Xu, J.; Wang, N.; Yu, D.P. Electrical Transport Measurements of the Side-Contacts and Embedded-End-Contacts of Platinum Leads on the Same SingleWalled Carbon Nanotube. Nanotechnology 2009, 20, 195202–195204. (51) Heinze, S.; Tersoff, J.; Martel, R.; Derycke, V.; Appenzeller, J.; Avouris, P. Carbon Nanotubes as Schottky Barrier Transistors. Phys. Rev. Lett. 2002, 89, 106801.
ACS Paragon Plus Environment
29
ACS Applied Materials & Interfaces
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 30 of 32
(52) Sze, S. M.; Ng, K. K. Physics of semiconductor devices.; John Wiley & sons: New York, 2006. (53) Wang, W.; Liu, Y.; Tang, L.; Jin, Y.; Zhao, T.; Xiu, F. Controllable Schottky Barriers between MoS2 and Permalloy. Scientific Reports 2014, 4, 6928. (54) Zhao, Y.; Li, D.; Xiao, L.; Liu, J.; Xiao, X.; Li, G.; Jin, Y.; Jiang, K.; Wang, J.; Fan, S.; Li, Q. Radiation Effects and Radiation Hardness Solutions for Single-Walled Carbon NanotubeBased Thin Film Transistors and Logic Devices. Carbon 2016, 108, 363–371. (55) Gong, C.; Colombo, L.; Wallace, R. M.; Cho, K. The Unusual Mechanism of Partial Fermi Level Pinning at Metal-MoS2 Interfaces. Nano Lett. 2014, 14, 1714–1720. (56) Kang, J.; Liu, W.; Sarkar, D.; Jena, D.; Banerjee, K. Computational Study of Metal Contacts to Monolayer Transition-Metal Dichalcogenide Semiconductors. Phys. Rev. X 2014, 4, 031005. (57) Moon, B. H.; Han, G. H.; Kim, H.; Choi, H.; Bae, J. J.; Kim, J.; Jin, Y.; Jeong, H. Y.; Joo, M.-K.; Lee, Y. H.; Lim, S. C. Junction-Structure-Dependent Schottky Barrier Inhomogeneity and Device Ideality of Monolayer MoS2 Field-Effect Transistors. ACS Appl. Mater. Interfaces 2017, 9, 11240−11246. (58) Su, J.; Feng, L. P.; Zhang, Y.; Liu, Z. T. The Modulation of Schottky Barriers of MetalMoS2 Contacts via BN-MoS2 Heterostructures. Phys. Chem. Chem. Phys. 2016, 18, 16882– 16889. (59) Tomer, D.; Rajput, S.; Li, L. Spatial Inhomogeneity in Schottky Barrier Height at Graphene/MoS2 Schottky Junctions. Journal of Physics D: Applied Physics. 2017, 50, 165301. (60) Si, J.; Liu, L.; Wang, F.; Zhang, Z.; Peng, L. Carbon Nanotube Self-Gating Diode and
ACS Paragon Plus Environment
30
Page 31 of 32
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
ACS Applied Materials & Interfaces
Application in Integrated Circuits. ACS Nano 2016, 10, 6737–6743. (61) Chen, B.; Zhang, P.; Ding, L.; Han, J.; Qiu, S.; Li, Q.; Zhang, Z.; Peng, L. Highly Uniform Carbon Nanotube Field-Effect Transistors and Medium Scale Integrated Circuits. Nano Lett. 2016, 16, 5120–5128. (62) Qiu, C.; Zhang, Z.; Xiao, M.; Yang, Y.; Zhong, D.; Peng, L. Scaling Carbon Nanotube Complementary Transistors to 5-nm Gate Lengths. Science 2017, 355, 271–276. (63) Ding, L.; Zhang, Z.; Liang, S.; Pei, T.; Wang, S.; Li, Y.; Zhou, W.; Liu, J.; Peng, L. CMOSBased Carbon Nanotube Pass-Transistor Logic Integrated Circuits. Nat. Comm. 2012, 3, 677–683. (64) Peng, N.; Li, H.; Zhang, Q. Nanoscale Contacts between Carbon Nanotubes and Metallic Pads. ACS Nano 2009, 3, 4117–4121. (65) Jeong, H.; Bang, S.; Oh, H. M.; Jeong, H. J.; An, S.; Han, G. H.; Kim, H.; Kim, K. K.; Park, J. C.; Lee, Y. H.; Lerondel, G.; Jeong, M. S. Semiconductor–Insulator–Semiconductor Diode Consisting of Monolayer MoS2, H-BN, and GaN Heterostructure. ACS Nano 2015, 9, 1003210038. (66) Zhao, Y.; Huo, Y.; Xiao, X.; Wang, Y.; Zhang, T.; Jiang, K.; Wang, J.; Fan, S.; Li, Q. Inverse Hysteresis and Ultrasmall Hysteresis Thin-Film Transistors Fabricated Using Sputtered Dielectrics. Adv. Electron. Mater. 2017, 3, 1600483.
ACS Paragon Plus Environment
31
ACS Applied Materials & Interfaces
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
373x285mm (144 x 144 DPI)
ACS Paragon Plus Environment
Page 32 of 32